Light Gluino Search for Decays Containing 7r + 7r or ir° from a
Neutral Hadron Beam at Fermilab.
A. Alavi-Harati 12 , I.F. Albuquerque 10 , T. Alexopoulos 12 , M. Arenton 11 , K. Arisaka 2 ,
S. Averitte 10 , A.R. Barker 5 , L. Bellantoni 7 , A. Bellavance 9 , J. Belz 10 , R. Ben-David 7 ,
D.R. Bergman 10 , E. Blucher 4 , G.J. Bock 7 , C. Bown 4 , S. Bright 4 , E. Cheu 1 , S. Childress 7 ,
R. Coleman 7 , M.D. Corcoran 9 , G. Corti 11 , B. Cox 11 , M.B. Crisler 7 , A.R. Erwin 12 ,
R. Ford 7 , A. Golossanov 11 , G. Graham 4 , J. Graham 4 , K. Hagan 11 , E. Halkiadakis 10 ,
K. Hanagaki 8 , S. Hidaka 8 , Y.B. Hsiung 7 , V. Jejer 11 , J. Jennings 2 , D.A. Jensen 7 ,
R. Kessler 4 , H.G.E. Kobrak 3 , J. LaDue 5 , A. Lath 10 -t, A. Ledovskoy 11 , P.L. McBride 7 ,
A.R McManus 11 , P. Mikelsons 5 , E. Monnier 4 -*, T. Nakaya 7 , U. Nauenberg 5 , K.S. Nelson 11 ,
H. Nguyen 7 , V. O'Dell 7 , M. Pang 7 , R. Pordes 7 , V. Prasad 4 , C. Qiao 4 , B. Quinn 4 ,
E.J. Ramberg 7 , R.E. Ray 7 , A. Roodman 4 , M. Sadamoto 8 , S. Schnetzer 10 , K. Senyo 8 ,
P. Shanahan 7 , P.S. Shawhan 4 , W. Slater 2 , N. Solomey 4 , S.V. Somalwar 10 , R.L. Stone 10 ,
I. Suzuki 8 , E.C. Swallow 4 ' 6 , R.A. Swanson 3 , S.A. Taegar 1 , R.J. Tesarek 10 , G.B. Thomson 10 ,
PA. Toale 5 , A.K. Tripathi 2 , R. Tschirhart 7 , Y.W. Wah 4 , J. Wang 1 , H.B. White 7 ,
J. Whitmore 7 , B. Winstein 4 , R. Winston 4 , J.-Y. Wu 5 , T. Yamanaka 8 , E.D. Ziimmerman
1 University of Arizona, Tucson, Arizona 85721
2 University of California at Los Angeles, Los Angeles, California 90095
3 University of California at San Diego, La Jolla, California 92093
The Enrico Fermi Institute, The University of Chicago, Chicago, Illinois 60637
5 University of Colorado, Boulder, Colorado 80309
6 Elmhurst College, Elmhurst, Illinois 60126
7 Fermi National Accelerator Laboratory, Batavia, Illinois 60510
Osaka University, Toyonaka, Osaka 560 Japan
9 Rice University, Houston, Texas 77005
' Rutgers University, Piscataway, New Jersey 08855
University of Virginia, Charlottesville, Virginia 22901
University of Wisconsin, Madison, Wisconsin 53706
We report on two null searches, one for the spontaneous appearance of ir + n~
pairs, another for a single 7r°, consistent with the decay of a long-lived neutral
particle into hadrons and an unseen neutral particle. For the lowest level
gluon-gluino bound state, known as the R°, we exclude the decays R° — >
7r + 7r~7 and R° — > 7r°7 for the masses of R° and 7 in the theoretically allowed
range. In the most interesting R° mass range, < 3 GeV/c 2 , we exclude
R° lifetimes from 3 x 1CT 10 seconds to as high as 10~ 3 seconds, assuming
perturbative QCD production for the R°.
PACS numbers: 13.85.Rm, 13.25.Es, 14.40.Aq, 14.80.Ly
Typeset using REVTpjX
Light masses for gluinos (g) and photinos (7) arise naturally in many supersymmetry
(SUSY) models, including those that solve the SUSY-CP problem by eliminating dimension-
3 operators [[J . They predict light gauginos and heavy squarks, and have not been ruled out
conclusively. In such models the gluino and photino masses are expected to be < 1.0 GeV/c 2 .
The gluino should form bound states with normal quarks and gluons (g), the lightest of which
is called the R°, a spin 1/2 gg bound state.
Estimates of the mass and lifetime of the R vary from 1 to 3 GeV/c and 10~ 10 to
1CT 5 s respectively [@]. Chung, Farrar, and Kolb |§ show that a stable 7 as a relic dark
matter candidate requires the ratio of masses M R o/M^ = r to be 1.3 < r < 1.8. The
range 1.3 < r < 1.55 is favored. Values of r below 1.3 yield an insufficient abundance of
dark matter, while too large a value of r overdoses the universe. A previous direct search
for the R° by the KTeV collaboration is described in That result, based on 5% of
the data collected by the KTeV experiment in 1996, excluded the R° with the constraint
M R0 - M7 > 0.648 GeV/c . Thus for r < 1.4, our previous result was insensitive to
M R o < 2.3 GeV/c 2 , which represents a large portion of the region of primary interest ||. A
search for the C-suppressed decay R° — > 777 is discussed in || . References to other searches
can be found in ||.
We assume exactly the same production mechanism and rates as described in ||. The
R° is expected to decay mainly into pj. The decay into 7r°7 or 777 is suppressed due to
conservation of C-parity. Figure |] illustrates the lower limit of sensitivity in M R o — My of
the three decay modes mentioned above, along with the cosmological constraints. We report
on null searches for two decay modes. The first is the dominant decay, R° — *> pry, p — > 7r + 7r~.
The second is the C-violating decay R° — > 7r°7. In both decays, the 7 escapes undetected.
The KTeV experiment as used in the R° search is described in |||. The data used in
the R° — > fry, p — > 7r + 7r _ analysis were collected during the 1996 run of KTeV (FNAL
E832). The trigger and analysis cuts used are similar to those described in To detect a
possible R° signal we examined decays with two charged particles; specifically the shape of
the invariant mass distribution, with the assumption that the particles were pions (M^^-).
An online filter was used during data collection to classify the events according to 7r + 7r~
invariant mass. The data with M n + W - < 0.45 GeV/c 2 were prescaled. Backgrounds consisted
of Kl — > 7r ± / =F z/ (/ = e, /x) decays with leptons mis-identified as pions (semi-leptonic decays);
Kl — ► 7T + 7r _ and if^ — > 7r + 7r _ 7 decays; as well as Kl — > 7r + 7r~7T° decays with undetected
The offline analysis, including cuts using photon veto energies, semileptonic decay rejec-
tion, and track, vetex quality requirements, was similar to that described in |4J]. Additional
cuts reduced the probability of track reconstruction errors that paired the wrong combina-
tion of horizontal and vertical track components. The Kl — > 7r + 7r~ decays were rejected
by requiring the square of the transverse momentum of the 7r + 7r~ with respect to the beam
direction (Pj) to be greater than 0.001 (GeV/c) 2 , and the Kl — > 7r + 7r~7 and Kl — > n + n^n
decays were rejected by using the calorimeter to identify the photons from respective decays.
In the region M n + n - < 0.36 GeV/c 2 additional cuts (Kl — > 7r + 7r~7r° specific cuts), including
restricting the total energy deposited in the calorimeter to < 5 GeV/c 2 , further reduced the
Kl —> 7i + 7T~7T° decays.
We performed a maximum-likelihood fit to the M n + n - distribution, using Monte Carlo
distributions for Kl — > vrez/, Kl — > 7r/iz/, — > 7r + 7r _ , — > 7r + 7r _ 7r°, and R° — > 7r + 7r _ 7.
The Monte Carlo events were subjected to all the same cuts as the data, however the e
identification cuts was not applied to the Kl —* vrez/, and a muon veto requirement was not
required for the Kl — > TTfiu events. These lepton identification cuts have no effect on the
Mn+ W - distribution. The amplitudes for all the simulated M w + W - shapes were allowed to
Figure 0(a) shows the M v + V - distribution for all data, before applying the Kl — > 7r + 7r~7r°
specific and P% cuts. There are ~ 2.1 xlO 6 CP-violating Kl — ► /T + 7r _ decays in the peak
at Mk- The sharp edge at M n + n - = 0.45 GeV/c 2 is due to the online filter prescale. The
Kl — > 7T + 7r _ 7r° decays are evident at M^,,— < 0.36 GeV/c 2 , and the kinematic limit is
evident at 2M n = 0.28 GeV/c . Also shown in Fi gure 0(a) is the sum of the various kaon
decay distributions from Monte Carlo. The data and kaon decay simulation are in agreement
for six orders of magnitude. In addition, the P£ distributions for data and sum of decay
simulations (not shown) are in good agreement, using the amplitudes for the various kaon
decays found in M n + n - fit.
Figure ^|(b) shows the M n + n - distribution for the data with all the cuts, as well as the
sum of the Kl decay simulations, and distributions for two sample -R°,7 combinations. The
Kl — > 7r + 7T - peak in data is significantly reduced due to the P% cut. The agreement between
data and the sum of Kl decay simulations has an overall x 2 /degree of freedom of ~ 194/148
for the region 0.28 GeV/c 2 < M n + n - < 0.58 GeV/c 2 .
The two sample R° distributions shown in figure ||(b), are scaled to the expected rate f7j
PI for R 0, s with a lifetime equal to the lifetime of the K L . Since the shape due to R° decay
is significantly different from those due to kaon decay, we searched for the R° by examining
the difference between the M n + n - shape and the shape expected from kaon decays. The data
show no deviation in the M n + V - shape that could indicate a contribution from an R° decay.
Limits on R° were obtained using the maximum likelihood fit explained above. There are
10 events with M 7r + n - > 0.6 GeV/c 2 that are not simulated by Kl decays. These events,
which are consistent with residual gas interactions in the vacuum, are treated as signal in
Various R°, 7 combinations, with 1.3 < r < 2.2 were used in the fit. All fits yielded
R° components consistent with zero. An upper limit for a given R° was determined by
evaluating the maximum-likelihood curve at the 90% confidence level (C.L.) interval, which
limited the R° decays to less than a few tens to hundreds range. The detector acceptance
and the R° flux at production were then determined as a function of the R° lifetime. The
normalization was performed using 2.1 xlO 6 Kl — > tt + tt~ events observed, from which we
determined that 37.7 xlO 10 K° exited the absorbers [|J. Figure |3| shows the 90% C.L. upper
limit on the ratio R°/K°, as well as the expectation for the R°/K° ratio for r = 1.4.
Figure || shows the variation of the 90% C.L. upper limit on the R°/K° flux ratio with the
R° lifetime for two sample R°, 7 combinations. Particles with lifetimes much shorter than the
Kl decay too close to the target to be visible in the detector, while those with much longer
lifetimes exit the detector without decaying. We use the R°/K° flux expectation to exclude
a range of lifetimes for a given R°, 7 combination. Figure [| shows R° lifetimes excluded at
90% CL. for a given mass, assuming a 100% branching fraction for the R° — > 7r + 7r~7 decay.
Contours are shown for r = 1.3, 1.4, 1.55, 1.73.
In this analysis, we are able to exclude _R°'s with masses well below the lower limits
of previous searches (~ 2.2 GeV/c 2 ). The lower limit of the exclusion contour (at 1.3
GeV/c 2 for r = 1.3) is due to the kinematic limit of Mro — My = 2M n . We note that
in the theoretically interesting regions of Mro and r, our exclusion covers lifetimes as low
as 3 x 10~ 10 seconds, and as high as 10~ 3 seconds, effectively spanning the theoretically
interesting range of lifetimes.
If the mass difference Mro — Mj is less than 2M V then the R° can only decay via
R° — > 7r°7. We searched for the decay R° —> 7r°7, tt° —> 77 in data taken during a special
run whose primary purpose was to search for the decay Kl —> 7i uV ||. Since the signatures
for the Kl and R° decays, two photons with missing transverse momentum (Pt), are similar,
these data are sensitive to -R — > 7r°7 decays.
Only one narrow beam of neutral kaons was used in this run, with the transverse beam
size of 4 cm x 4 cm at the calorimeter. The trigger was designed to select events with
two energy clusters in the calorimeter, together with four cluster events (Kl — > 7r°7r°) for
normalization. The longitudinal distance of the decay vertex from the target (Z) and the
transverse momentum of the two photons were determined by constraining the invariant
mass of the two photons to that of tt°. The average Pt resolution was ~ 8 MeV/c. The
selection criteria used in this data sample are similar to the one used in the ir°vT> analysis 0,
with the exception of the Pt cut at 260 MeV/c. Photon veto detectors and drift chambers
were used to suppress backgrounds from other kaon decays and hadronic interactions in the
detector. The events were required to have the decay vertex in vacuum, with the vertex Z
position in the range 125 < Z < 157 m.
We examined the shape of the Pt distribution to isolate R° candidates. The Pt distri-
bution of the final data sample is shown in Figure ||, along with the background expected
from Kl — > 77 and A — > mr° decays. The peak near Pt = is from the decay Kl — > 77,
and the remaining events below Pt = 160 MeV/c are from A decays. The signal search
region at Pt > 160 MeV/c, chosen to minimize background from hyperon and Kl decays,
is indicated by the arrow. Clearly, we are sensitive to R° masses for which the 7r + 7r~ decay
After all cuts, two events remain in the signal region. From studies of A — > nir° decays |§ ,
we expect the number of events from hadronic interactions to be 4.7 ± 1, consistent with
the number of events remaining in our sample. Treating these two events as signal, the
corresponding 90% CL. upper limit on the number of observed R° signal events is 5.32.
The Kt flux in this data sample was measured from 3466 observed Kl — > 7r°7r° decays.
The pQCD prediction for the R°/K L flux ratio were used as before to obtain upper and
lower lifetime limits at the 90% C.L., assuming the R° decays 100% of the time to 7r 0i y.
Figure |^ shows the exclusion contours for r = 1.3, and r = 1.4 using the 7r° analysis, and
the 7T + 7i~ analysis. Note that using the 7r° analysis, we extend the range of excluded M R o
down to 0.8 GeV/c 2 , for lifetimes between 2.5 x 10~ 10 and 5.6 x 10 -6 seconds.
The analyses presented in this paper exclude most R° masses, over six decades in lifetime.
A significant portion of the region allowed by the cosmological constraint r < 1.4 and
M R o < 2.2 GeV/c which was not addressed by previous searches is now excluded. We thus
definitively close the light gaugino window. Our null results eliminate most SUSY models
in which gauginos remain massless at tree-level. More generally, our understanding of the
M n + n - shape will constrain future models that predict long-lived particles with a 7r + 7r~
component in their decays.
We thank Glennys Farrar for suggesting this search and for discussions concerning this
work and, along with Rocky Kolb, for pointing out the cosmological significance of this
search. We gratefully acknowledge the support and effort of the Fermilab staff and the
technical staffs of the participating institutions for their vital contributions. This work was
supported in part by the U.S. DOE, The National Science Foundation and The Ministry of
Education and Science of Japan. In addition, A.R.B., E.B. and S.V.S. acknowledge support
from the NYI program of the NSF; A.R.B. and E.B. from the Alfred P. Sloan Foundation;
E.B. from the OJI program of the DOE; K.H., T.N. and M.S. from the Japan Society for
the Promotion of Science. P.S.S. acknowledges receipt of a Grainger Fellowship.
t To whom correspondence should be addressed.
Electronic address: firstname.lastname@example.org
* On leave from C.P.P. Marseille/C.N.R.S., France.
 R. Mohapatra and S. Nandi, Phys. Rev. Lett. 79, 181 (1997); Z. Chacko et al, Phys.
Rev. D56, 5466 (1997); S. Raby, Phys. Rev. D56, 2852 (1997).
 G.R. Farrar, Phys. Rev. Lett. 76, 4111 (1996); G.R. Farrar, Phys. Rev. D 51, 3904
 D.J.H. Chung, G.R. Farrar, and E.W. Kolb, Phys Rev. D 56, 6096 (1997); G.R. Farrar
and E.W. Kolb, Phys. Rev. D 53, 2990 (1996).
 J. Adams et al, Phys Rev. Lett 79, 4083 (1997).
 G. R. Farrar, Nucl. Phys. Proc. Suppl. 62, 485 (1998) (available as eprint [hcp-
| ph/9710277| ).
 V. Fanti et al. Phys. Lett. B 446, 117 (1999).
 S. Dawson, E. Eichten, and C. Quigg, Phys. Rev. D 31, 1581 (1985).
 C. Quigg (private communication).
 J. Adams et al, Phys. Lett. B 447, 240 (1999).
- J? -7*jp$£ y (Dominant Decay Mode)
- 1 i i i i 1 i i i i 1 i i i i 1 i i i i 1 i i i i 1 i i i i 1 i i i i 1 i i i i 1 i i i i 1 i i
1.25 1.5 1.75
2.25 2.5 2.75 7
M R0 (GeV/c 2 )
FIG. 1. M R o — vs. M R o, showing the region allowed by cosmological arguments,
1.3 < r < 1.8 (hatched). The dashed lines represent the lowest level of sensitivity for the var-
ious decay modes.
i 10 5
a) All P t 2
b) P t 2 > 0.001
Sum of All Kaon Decay
Monte Carlo Components
- - Kj—^izev Component
"" Kj^iz\m Component
II I J
M R0 M~
FIG. 2. M^+ n - distribution with all but P% and — > 7r + 7r~7r specific cuts (a), and all cuts
(b). The data (dots), sum of all MC (solid line) are shown. Suppression below 0.45 GeV/c 2 is
due to the online filter prescale. The separate contributions from semileptonic decays (dashed and
dotted lines) are also shown in (a). Distributions due to two sample R 0, s of lifetime 5 x 10~ 8 sec
using the predicted flux (dashed and dotted lines) are also shown in (b).
1(1 I I I I I I I I I I I I I I I I I I l_
v -9 -8 -7 -6 -5
10 10 10 10 Q 10
R Lifetime (sec)
FIG. 4. Upper limits with 90% C.L. on the R°/K° flux ratio as a function of R° lifetime, for
two M R o, My combinations, with masses listed in GeV/c 2 . The dotted lines show the expectation
for the flux ratio, and the stars mark the corresponding lifetime limits.
/ r _ J 3
) r = 1.4
i 1 i 1 i 1 i 1 i 1 i 1 i 1
i i i i i i i i i i i i i i
) r = 1.73
i i i i i i i i i i i i i i
i i i i i i i i i i i i i i
-9 -7 -5
10 10 10
R Lifetime (sec)
-7 -5 -3
10 10 10
R Lifetime (sec)
FIG. 5. R° mass-lifetime regions excluded at 90% C.L., for the decay R° — > tt + tt 7, for values
of r = 1.3, 1.4, 1.55, and 1.73. The lower edges are due to the kinematic limit of M R o — My = 2M W .
P T (MeV/c)
FIG. 6. The Pt distribution for 77 events from the narrow-beam run (dots). The background
from Kl and hyperon decays is shown by the solid line. The arrow indicates the R° signal region.
The MC simulation for M R o = 0.8, l.OGeV/c (dashed, dotted lines) are also shown, both for
r = 1.4.